Fueling for Ultras Is an Enigma

The sport of ultrarunning was essentially nonexistent 50 years ago. Even as the running boom took hold in the 1970s, races longer than the standard 26.2-mile marathon remained a rarity. The origins of perhaps the most famed ultrarace in the United States, the Western States 100-mile Endurance Run (WSER), date back to 1974 when Gordy Ansleigh ran on foot against horseback riders in the Western States Trail Ride, but the first official WSER wasn’t held until 1977.

The popularity of ultrarunning has grown dramatically in the intervening years, particularly in the past two decades. According to Ultarrunning.com, there were over 2,400 ultrarunning races in North America in 2019, with 128,685 finishes. In contrast, there were just over 200 such races and 13,852 finishes in 2000. These are still relatively small counts compared to shorter running races, but there’s no reason to think that the number of people competing in ultras won’t continue to grow precipitously in the coming decades.

Because ultrarunning places such extreme energy and metabolic demands on the body, fueling is undoubtedly one of the most important elements to crossing the finish line of these races. Paradoxically, however, the extreme durations involved in these contests also make them a particularly tough nut for scientists to crack. And although this is changing ever so slowly, to a large degree we scientists are akin to a person wandering around in the dark looking for a light switch. And ultimately, this lack of data explains why personal anecdotes and trial and error play such a huge role in informing nutrition practices in ultrarunning communities.

Why the Lack of Data?

From the simplest of perspectives, studying nutrition in ultrarunning is so vexing because of the exercise durations involved. As a scientist, one of my major concerns when designing sports nutrition experiments is participant recruitment. If you can’t get enough people for a good-sized sample, there’s no point in even carrying out your experiment, as you won’t be able to say whether any changes you see are due to the intervention or simply from day-to-day variation in performance. The definition of an adequate sample varies depending on the scientific field and a study’s methodology, but any experiment with less than 10 people is usually highly questionable.

In most metropolitan areas, it’s not hard to find a few dozen committed ultrarunners, if not more. Convincing them to volunteer hours of their time, often for no money, while being poked and prodded like a lab rat, is another matter. The best studies also ask participants to standardize their diet and training, meaning that these human guinea pigs must give up control over these things for at least a few weeks. It’s safe to say most runners are not big fans of relinquishing control over these particulars.

To put these issues into more concrete terms, it can be helpful to consider a hypothetical example. Say I wanted to understand how two different in-race fueling strategies impact 50-mile ultra performance. Finishing times for 50-mile ultras vary a lot depending on the course, but assume I design a protocol on a treadmill that takes eight hours to finish on average. Most of these types of studies ask participants to complete all the study interventions over a period of weeks or months, which effectively means that each person serves as their own control. In addition, each person would ideally complete a baseline performance test prior to receiving any intervention. To recap, that means each volunteer is expected to complete three separate 50-mile runs on a treadmill. Not only that, but if you want to understand whether an intervention actually impacts performance, your volunteers need to give a good effort each time.

Even if a researcher somehow miraculously found enough volunteers to do this sort of study, there are more issues to consider. As any ultrarunner could tell you, replicating all the facets of a real-life race is next to impossible in a lab (changes in elevation, the isolation on certain parts of the course, the extremes in environmental temperature, etc.). And that begs the question as to how well any result from a lab study translates to the real world. A seemingly simple solution to this problem would be to do an experiment at an actual ultrarace, but perhaps unsurprisingly, runners often don’t want to risk volunteering for a study where they must stick to a rigid fueling plan they may be unaccustomed to. Furthermore, the optimal fueling plan for an elite runner who competes for a win is inherently different from someone running near the middle or back of the pack.

Long story short, there are few published experiments that have scrutinized the effects of various nutrition interventions on ultra-exercise performance. This was succinctly summarized by the authors of a review paper on single-stage ultramarathoning that was recently published by the International Society of Sports Nutrition.

“The data informing our recommendations are incomplete . . . for several reasons. Firstly, despite the growing popularity of ultra-marathon, participant numbers are still relatively low. Moreover, runners are often reluctant to compromise their race preparation and/or recovery to volunteer for data-collection . . . ”

What’s more, Dr. Martin Hoffman, who has served as the research director for the WSER and has led numerous ultrarunning research studies, recently combed the scientific literature and found that, from 1999 to 2019, there were only 1.6 experimental-type studies done per year on the topic of ultrarunning.

So How Do Most Ultrarunners Fuel?

Despite the lack of experimental data on ultramarathon fueling, we do have a decent amount of research available on what these athletes choose to do of their own volition. One of the earliest published accounts was an examination of what the Greek ultrarunner Yiannis Kouros ate during the 1985 Sydney–Melbourne Ultramarathon. Kouros is widely considered as one of the greatest ultrarunners of all time and has sometimes been called the “Running God” (he even starred as Pheidippides in the movie The Story of the Marathon: A Hero’s Journey). During his 5-day, 5-hour Sydney–Melbourne Ultramarathon effort, Kouros downed just over 55,000 kcal, or about 10,500 kcal per day. The vast majority of his energy intake (96 percent) came from carbohydrate (Greek sweets, dried fruit, fruit and sport drinks, etc.).

It’s certainly easier to maintain energy balance during multi-day ultramarathons like the one Kouros did in 1985, so we need to examine the fueling practices of single-stage races through a different lens. Several organizations such as the American College of Sports Medicine say that athletes participating in ultra events that last more than 2.5 hours can consume up to 90 grams of carbohydrate per hour, whereas intake rates of 30–60 grams per hour are more realistic for shorter events. The difference in recommendations is largely because it is easier to consume more food and carbohydrate during longer events (at least in theory), as the average exercise intensity is lower.

So, where do ultrarunners tend to fall when it comes to carbohydrate intakes? Well, most eat nowhere near 90 grams of carbohydrate per hour. Instead, many studies find average ingestion rates closer to 30 to 40 grams per hour (see here and here). The reasons for these lowish intake rates are varied. Many of the runners studied were average to back-of-the-pack finishers, who likely needed to eat less carbohydrate on a per hour basis than highly competitive runners. In addition, gut distress (including nausea) is very common in ultraraces and can interfere with a runner’s ability to tolerate eating and drinking. In one study, for instance, intakes of energy and carbohydrate were negatively correlated with several gut complaints (vomiting, heartburn, diarrhea) during a 60-kilometer ultramarathon.

Although most investigations find that ultrarunners tend to eat 20 to 40 grams of carbohydrate per hour during competition, there is some interesting evidence that higher intake rates can be tolerated by certain athletes. One of the studies I referenced earlier included a runner who consumed 108 grams of carbohydrate per hour during a 120-kilometer race. Another investigation of athletes from Ironman races found that a handful of them ate about 120 grams of carbohydrate per hour, which is equivalent to 5 to 6 sports gels every hour! A group of Spanish researchers recently took these field observations even further and found that most runners were able tolerate these high intake rates relatively well during a mountain marathon so long as they trained their guts before competition.

As a caveat, it’s important to note that these studies can’t tell us much about carbohydrate dosage and performance. To put it another way, does downing 120 grams of carbohydrate per hour—if tolerated well by the gut—lead to better performance than a more modest intake (30–60 grams per hour)? Unfortunately, we do not yet have good experimental evidence to answer this question, at least when it comes to ultra racing. And honestly, the answer will likely depend on a number of factors, including the specific race, individual differences in metabolism, and whether the athlete has trained their gut to tolerate high ingestion rates.

Formulating Fueling Recommendations in the Face of Slim Evidence

Given that the cupboard of science is pretty bare when it comes to ultrarunning and fueling, recommendations from nutritionists and dietitians tend to focus on individualism, athlete experience, and trial and error. To illustrate this point further, consider that the recent position stand from the International Society of Sports Nutrition on nutritional considerations for single-stage ultramarathoning included references to individualization and personal tolerance more than a dozen times. In other words, there are no quick and easy answers.

While we wait (hopefully not in vain) for more and better experimental science to emerge on how to fuel for an ultra, I would urge athletes to regularly document their feelings, perceptions, and challenges with fueling during training and racing. This doesn’t have to be a high-tech endeavor; paper and pen will suffice in most cases. The key is to keep some record of how your various nutrition plan adjustments and iterations affect how you feel, recover, and perform. Ultimately, this sort of regimented approach may help speed up the learning process that comes naturally with experience over time.